Microdata files are now available to PC users on CD-ROM disks that hold the equivalent of more than four reels of mainframe magnetic tape, but CD-ROM drives have a very slow read cycle. In Phase I, the main goal is to learn how to overcome this by using innovative software techniques. Investigators will study commercial and public domain software systems to see if one or more of them could be modified to access CD-ROM based microdata files quickly and efficiently. IDD will create a partial prototype system that will be used to develop baseline data on access times using different access methods. This data will form the basis for the design of a CD-ROM microdata processing system for Phase II implementation. The Phase I target test files will be selected from some of the census Public Use Microdata Samples (PUMS). PUMS will soon exist for most decades since 1880. In addition there are many other large microdata file series that are important to investigators in demography, epidemiology, health care management, and other health-related fields. Work in Phase II will develop a library of CD-ROM microdata volumes for sale as a Phase III commercial venture.